Bilingual AI Safety Data Evaluator (English/Spanish)

About OpenTrain

OpenTrain is the #1 platform for building careers in AI training and data labeling. We connect people to meaningful work that directly shapes how modern AI systems behave and offers flexible, remote opportunities across projects and specialties.

As the hiring and contracting organization for this role, OpenTrain provides the project framework, training materials, and pay. Creating an OpenTrain account is free and is the first step to apply.

About AI training and this kind of work

AI training (also called data labeling or human feedback work) is the human side of building artificial intelligence — people prepare and review examples that teach models how to respond safely and usefully. This role focuses on safety and reasoning evaluation for large language models, a high-impact area where human judgment shapes model behavior.

Work in AI training is often remote and can fit part-time schedules; many contributors use it to gain experience in the cutting-edge field of model safety and content moderation.

The role

You will review AI-generated text in English and Spanish and provide expert annotations and feedback focused on safety, accuracy, reasoning, and policy consistency. The role is contract-based, hourly paid, and remote (worldwide).

Tasks include labeling and rating model outputs, quality-checking safety data, red-teaming edge cases, and applying nuanced policy judgments to detect risky, harmful, or otherwise unsafe responses.

What you'll do

• Evaluate and rate AI-generated responses for safety, factuality, logic, and clarity across English and Spanish content.
• Apply policy guidelines to identify hate, harassment, sexual content, self-harm, violence, bias, illegal goods/services, malicious activities, malicious code, and misinformation.
• Perform adversarial testing and red-teaming to surface edge cases and recommend mitigation strategies.
• Create clear, reproducible rationales for moderation and safety decisions used to train and improve models.
• Quality-check annotations and contribute to documenting ambiguous or culturally sensitive cases.

Minimum qualifications

• Near-native or native Spanish proficiency in reading and writing.
• Minimum C1 English proficiency in reading and writing.
• Bachelor’s degree or higher in Communications, Linguistics, Psychology, Law/Policy, Security Studies, or equivalent professional experience.
• 5+ years professional experience in Trust & Safety, content moderation, policy operations, risk, compliance, investigations, or related safety work.
• Proven LLM red-teaming or adversarial testing experience, including identifying edge cases and recommending mitigations.
• Strong knowledge of safety domains listed above and experience applying policy consistently across multilingual or cross-cultural content.
• Comfortable reviewing explicit, toxic, violent, sexual, or psychologically disturbing content as part of daily work.

Preferred skills and experience

Localization or translation experience is preferred, especially the ability to preserve meaning, severity, and intent across Spanish and English.

Strong analytical writing skills and the ability to produce concise, reproducible rationales for safety and moderation decisions are highly valued.

• Experience working with multilingual policy frameworks or moderation guidelines.
• Prior hands-on experience in RLHF, evaluation rating, or text generation assessment workflows.

Compensation, schedule, and logistics

This is a remote contractor role, open worldwide. Employment type: Contractor.

Pay is hourly: listed rate $20/hr with a range of $14–$24/hr. Exact schedule and hours will be defined when you are onboarded to the project.

Work is text-based (evaluation, RLHF, text generation). You may encounter sensitive or disturbing content and should have the emotional resilience to handle that material.

Who should apply

Apply if you are an experienced trust & safety or moderation professional with strong bilingual Spanish/English skills, proven LLM red-teaming experience, and a track record applying policy consistently across languages or cultures.

This role suits candidates who enjoy analytical writing, making defensible safety judgments, and helping shape how AI systems handle risky content.

How to apply

Create (or sign in to) your free OpenTrain account, complete your profile, and submit your application for this role. Include examples of relevant experience and any localization or red-teaming work you’ve done.

Qualified applicants will be contacted with next steps. OpenTrain provides the project framework, training materials, and payment for successful contributors.